The primary hypothesis to be tested in this project is when milk allergic individuals undergo oral immunotherapy (OIT), the induction of milk-specific regulatory T cells will predict those individuals who achieve the most complete and lasting clinical tolerance. In order to test this hypothesis, we propose to recruit 60 milk allergic children and adults (>15 years) who are very unlikely to outgrow their milk allergy without an intervention. These individuals will be randomized in a 2:1 ratio to receive either active treatment with milk OIT or placebo treatment. They will undergo three oral challenges to milk (the placebo group will have only two challenges) to define their clinical sensitivity before treatment, after reaching maintenance for about three months, and - if they become less sensitive - again after a three month period of avoidance. In this way, we expect to define groups who: 1. fail to improve or are unable to reach maintenance; 2. Improve but relapse after avoidance; 3. Achieve more lasting tolerance. We will compare these clinical outcomes to changes in the frequency and function of CD4 T cell subsets, including those that produce cytokines (T effectors), and those that are regulatory - defined by their expression of specific proteins (CD25, FoxP3) and not others (CD127) - as well as by their functional capacity to suppress other T cells. We will also study global gene expression changes in T cell subsets during OIT to investigate other potential mechanisms of OIT. We will attempt to test the hypothesis that an expansion in numbers of a specific T cell subset, such as the regulatory subset, occurs as a result of incorporating a greater diversity of T cells into that subset and examine the kinship between different T cell subsets. Finally, we will invest in an effort to define the specific regions of two major milk allergens that are recognized by human T cells and develop highly specific reagents (class II tetramers) that will advance the study of milk allergen specific T cells in any immunological disease for which it might be useful.